 Welcome to the best panel in Davos today. We're going to be talking about strategic outlook on the digital economy. I'm joined by an absolutely extraordinary group of people. They have, I think, a million and a half employees in total. Their companies are worth a combined about $2 trillion. They are all incredibly smart. I've spent a ton of time listening to them all and podcasting on YouTube at 2x speed, so I have no idea what they actually sound like in real life. But they all say brilliant, interesting things. We're going to begin by discussing a little bit about what we learned during the pandemic. We're going to then discuss a little bit about where we are. And we're going to discuss a little bit about where we're going. So starting off, we have Ruth Pratt. She's the CFO of Google. We have Arvind Krishna. He's the CEO of IBM. We have Pekka Lundmark. He is the CEO of Nokia. We have Antonio Neri. He is the CEO of HPE. And we have Julie Sweet. And she is the CEO of Accenture. Let's get cracking, Julie. So in my little Brooklyn media filter bubble, there are three issues that can blow up a dinner table conversation. First is transgender women in sports. Second is Web 3. And the third is return to office. Bring any of those up, and people just start fighting, even if they agree on everything else. So you've done a lot of research on return to office, and you found some crazy, interesting things. So tell me what you found in your research, and let's figure out whether we can get this right. Great. Well, first of all, I had 800 leaders together for the first time since the pandemic two weeks ago. And we did not have one topic about return to office. And why was that? Because what we found is the most important thing is about how people feel, whether they're in the office or not. So we have a leadership essential about caring for people personally and professionally and for the first time in two years, we got together. And the first topic was that topic. We had Michael Bush from The Great Place to Work and Michael Phelps talking about mental wellness and peak performance. And so the first thing I would say is we need to move away from talking about being in the office or not in the office and really focusing on people. And the research we released actually today would tell you, so it's recent research, that only one in six employees really feel connected. And the scores were not better for those who were returning to the workplace. And so I think that should be a lesson that it's not about spaces and places, but how you are connecting. We're the scores worse. The people who are coming to the office feel less connected. That's what we found. Now, that was very controversial because everyone said, you know, wait a minute, that's not what we're hearing. But this is research that's not anecdotal and it's persistent, right? So we have a philosophy at Accenture and what we're helping our clients do and it's called earn their commute. And so my guess is for those companies who are thinking about when they bring people back is there a purpose, right? Have they earned the commute? Those people who are probably feeling more connected, right? So is that convincing? You guys are all going to send your employees home, shut down Google's new beautiful offices? I actually think the point of earn the commute is the right way to think about it. So when we look at return to office, the keyword we're using is experimentation because we think it's too early to know after two years of something that's never been done before, what people really want. And we're experimenting on what's the structure of the week. We're experimenting on how we use space. Space has to be about bringing people together and collaborating so we have walls that move and you can convene in different ways and we're integrating Google Workspace and technology. And it's about being deliberate about how you lead. So it's not about being in the office to just to be in the office, it's about being there for a purpose. And so I do think it's too early to judge but we do need to experiment. And I would also underscore Julie's point high on that list is making sure we continue to focus on wellness because it is continuing to be a real tax on all of our people and in particulars are thinking about how do they manage children? And so focusing on wellness and wellness benefits we think is key. Well, I think return to Davos has totally been worth it much better than the last Zoom Davos. So I'm down with that. Antonio, let's talk about, you wrote a piece for the forum and everybody probably read it but appeared yesterday. And in it, you said something that I don't understand but sounded awesome. You said that companies should start putting data on their balance sheets. What did you mean or that they will? Maybe not that they should but that they will. What does that mean? Well, it means that today data is the most precious asset you have. Think about the digital economy we all live and operate. What we all realize, we are entering the new age of insight. The new age of insight is how we use data to advance societal challenges. It's not just the business. And we as a leadership of amazing companies, you have the responsibility to work with the public sector to figure out how to use that data to solve some of the societal problems we all live. Climate, inclusion, diversity, connectivity, so to me, obviously cybersecurity goes with it. So to me, it is a big opportunity. It's a turning point in our history where data is growing at a pace we had never seen before. However, we have to move faster and extract it in sites. And I believe at some point, as I said, that data will have to be recorded on the balance sheet. Not different than any other asset you have in your company, buildings and other intangible assets. So that's why it's so important we focus on that data. Well, explain a little bit the process of how we get from here to there. How would you actually, tell me a little bit more about how you would value your data. A building you can figure out, there's a market and it's building, it's worth this. How do you figure out your data? Well, I think if you think about it for a moment, right? So we use data for different means, whether it's to improve cybersecurity, to improve revenue streams, to improve experiences. We talk about the future of work, right? So it's all about providing digital experience to employees to do their job in an integrator way in a culture of the fabric of the culture. So what's the thing about the value of data? If it has a return, that insight has a return to improve revenue or to reduce cost, eventually you will be able to calculate how much that return is, and then eventually record that in your balance sheet. Not different than any other thing you put, my building is worth $200 million, great. Well now, that data has a tremendous value. Why don't I put it on the balance sheet? You guys all run balance sheets, you gonna do this? She's the CFO, you look skeptical. I think it's an intriguing concept. I think at the core of what you're saying, I would agree, which is using data wisely and really accessing the analytical tools that are available is absolutely critical if we wanna outperform. And one of the things that we're excited about is the fact that there have been these breakthroughs in the way we're looking at things. For example, Google Translate, one of the biggest breakthroughs we announced at our most recent developers conference, where we now don't need the full corpus of data that we previously had in natural language translation. So we just announced that we have introduced English Bengali translation using very limited data. And this type of move says that we're gonna be moving more rapidly with the limited data that you have. We just announced that we are going to have 24 new languages, 300 million people will have the ability to have natural language translation, Google Translate. Those kinds of things are exciting because what it says is start with data and you can continue to build on it and get ever-increasing insights. That's interesting, so it suggests if data is on a balance sheet, having less data will become more valuable in the future, data itself will become more valuable because you'll be able to do more with less. But also the value of the data will depend on the intelligence of the company because they're not an intelligent company with a lot of data, you can't really value that much. Real estate's the same for everybody because you just put it out in the market. Data is what you can pull out of it, is that the right way to think of it? Yeah, but listen, storing the amount of data we have today costs money, right? You have to store it, you have to make a compliance, you have to bring governance around it, hopefully create data pipelines and eventually run the algorithm that Ruth just talked about it, right? There is a whole value chain there that eventually need to understand the cost and the opportunity to generate return from that data. And not different than we do in other parts of the business. All right, Pekka, I'm gonna move to you. And Pekka is in Davos, and he says every single conversation and every bar he's been to, he's been asked about the Russian border with Finland. And so that is the only thing that is off limits in this conversation. We are not gonna talk about Russia and Finland. So instead we're gonna talk about the industrial metaverse. We are extremely excited about the industrial metaverse. What is it, why are you excited about it and how will you help block the Russian invasion of Finland with the industrial metaverse? Yeah. Thank you for avoiding the subject. I would actually like to build on what Antonio just said. I mean, you said that we should put data on our balance sheets. In a way, I agree. I need to discuss with my CFO how we do that in practice. But the point is that only a small fraction of industrial data in the world today is utilized. I mean, the industrial machines, the CNC machines, turbines, generators, you name it, robots, they produce massive amounts of data and we are not utilizing the data. So we need to connect these machines. We need to crunch the data and make sense of it with the help of AI in the cloud. And why is this important? There is no green without digital. We need to connect the digital agenda and the green agenda. And by using the data, we will be able to significantly improve the productivity and sustainability of our industrial processes in the world, which are the key reasons for the sustainability problems that we have, but not only sustainability, also equal access to work, equal access to opportunity, everything, only 30% of the world economy is digitalized today. And if we want to take care of this world, we need to digitalize the rest of the 70%. So I'm going to go back to the industrial metaverse in a second, because I do want you to define it and explain it, but first I'm going to ask Arvin to comment on that, explain your hypothesis for how we can best use data, best use the digital economy to help us meet our climate goals. And then we're going back to the industrial metaverse because I'm not going to let you out without explaining it. Arvin. Look, when you look at climate data, first of all, we are all obsessed, at least from the media and the policy side on transportation. But transportation is about 14%, 16% of all of pollution and true sustainability is going to come from cement, steel, agriculture, industrial goods. And so we got to attack that whole problem. So going back to what Pekka just said and what Ruth was implying, you got to collect the data across all these things to begin to understand what are you going to do? And it's only when you attack the whole problem are we truly going to make an impact on sustainability? Should we go more circular? What are the innovations needed to be able to capture carbon? Should we get more efficient? We know that 30% inefficiency exists upstream and there's another 30% downstream. We can go capture those. And as we begin to look at the data, we'll be able to decide which of those problems is the most attackable in the near term and which one needs more innovation and more invention in the medium term. And that is why to me, the data underlying all these is what is going to go make the difference to how we begin to attack these problems. So tell me something that will surprise me, anyone here, about how data has been used to reduce environmental impact or to reduce CO2 emissions. Tell me something that you've seen. You said, oh, that's interesting. We'll do this. Oh, really, very simple. So we worked with the bank in Australia. We just put simple sensors into all of the branches and you'll be shocked. 30% of the time the air conditioning is running full blast in the middle of the night with no people, lights were on all the time. You can go save 30% of the energy consumption because you shut it off at night and that's an and. You save money and you're good for the environment. Really simple example. This is why there is so much low hanging fruit in data. Forget the more deep stuff, which will even add more. Why do we flare methane off upstream? Why don't we collect it and give it? There's a shortage of natural gas in the world. Two simple examples. Methane sensors cost a dollar each. Anybody else want to jump in on this question? I know it's dear to all of your hearts, Ruth. I would say one of the most valuable assets that we have is geospatial data because when you combine geospatial data with data analytics and AI, it opens an understanding that can yield all sorts of results. So for example, we worked with UPS. They wanted to understand how to optimize the routes for their fleet of trucks. The first thing that we uncovered is we could save them 400 million a year by getting more efficient routes. But what does that do? That also reduces fuel consumption. We're working with Unilever. They wanted to understand what's going on with deforestation through their entire supply chain. And again, with geospatial data, data analytics and AI, we're helping them look through the supply chain and understand what's going on deforestation with one of the most important ingredient products, Palm. And so this geospatial data, we're also using it with cities. We're giving them an overview of what's their carbon footprint from the built environment in transit. Geospatial data for every one of us can help us understand what we're doing and how to optimize. Nick, I think what's important is this is a massive paradigm shift. This discussion blows up the idea that you have trade-offs between good business and societal benefits. And even this panel still talked about trade-offs. It's really about synergies. And so what's the biggest hope that we can actually change on sustainability? It's the fact that IT spending today is 5% of the GDP. And in 2030, it'll be nearly 7%. And so if we actually change the ways of working so that the same people that are making decisions on, we need to cut costs in our logistics, and those that understand sustainability are working together, that when you're digitizing manufacturing, you've got the expertise in power as well as automation put together. And that's not happening at many companies today because they're siloed. Those thoughts are different. And that's the role that we have to play. And I think the World Economic Forum is playing and sort of bringing these things together. But we're really moving from trade-offs to synergies. And that is a big shift. Let's talk a little bit about some of the missing data. That's super interesting. I want to get back to trade-offs. I was struck thinking about data as you were all talking, that there's all this data in the world. We should definitely share it. We can learn all these things. It's really important. You guys are all doing a great job at it. But there's one giant thing missing, which is China, right? Is there anybody from China in this room? We're here at the World Economic Forum and we're missing a big part of the world. We've got all the world's data, but Google isn't really in China. How do we actually make progress using data to solve these problems without any Chinese data or tech integration in China? That's a simple, easy question. Anybody want to take that? This is where we all say, not it. Arrant, you are eager to go. Look, if I look at most of these problems, if you have about 80% of the data, you can go a long way in terms of solving the problems. Now, are you going to implement it in a given country? Now, you get into all of the regulations and all of the politics. But that should not stop us from trying to solve the problem. If you solve carbon sequestration in the United States or in Brazil, I'm willing to bet the same science works in China. So I'd rather make that bet and then be an optimist that says if people are going to go see a benefit, then they're willing to go do that. And then to say, OK, I got to wait for everybody to come along for the ride. That's never going to happen. But wouldn't you be able to do a better job and a quicker job of solving carbon sequestration if you had data on how carbon sequestration works that was shared across the entire world instead of missing a huge part of it? I think that if you generally have 60%, 70%, you can get almost all the way there. It's not that you need 100% of the data to do something. I mean, to go back to Ruth's example on English, being all in with a little bit of data, the fact is there's a lot of commonalities. So would those scientists help? Help? Sure. But I think it'll be a 5% acceleration, not a 95% acceleration. All right, well, let's stay on China because it seems like a fun topic that puts everybody at ease. Who here has it? Do any of you do significant work in China? Yeah, we do. And tell me the hardest part about it, the most interesting part about it and what you've learned. Because I think tech companies should do a lot more. The world would be much better if we could get the tech industry fully integrated. So what do we need to do to get the Western tech industry integrated with the Chinese tech industry? Antonio, you're smiling. Well, we have a unique setup in China which is a partnership of sorts that we decided in 2015 that it would be better served to work in China with a Chinese champion than being a multinational in China. But that's more a structure that allows us to participate in the second largest market and probably the fastest growing market on the IT side. However, going back to the comment, I think we confuse concepts in my mind. So, sharing data. I think we should not talk about sharing data. We should talk about sharing insights. You know, there's a concept of, obviously, Hill-Ruth understand these and all of us in the context of data sovereignty, which obviously the public sector has to have a managed aspect on the policy side and regulation side, although the private sector has the responsibility to work with them because innovation is moving to a pace that the public sector is not able to keep up. So, but on the other side, it's how we start with a purpose, whether it's climate, whether it's food, whether it is, you know, healthcare, and then share insights. If you think about autonomous driving for a moment, right? So everybody's developing self-driving cars and assisted driving cars and so forth. You will know this will be proprietary stacks that everybody wants to protect. Is there data? I collect this and the data from the sensor and I have that data. But if you, Nick, you're driving ahead of me, take out a brand and I'll drive behind you why you will now share the insight ahead of you. There is a pothole and there is an eyes of sort, black eyes of sort. You can still monetize that insight, but you still keep the data. And I think that's the concept that we have to explore going forward because the world is becoming hyper-connected. More data is gonna create it where we live, where we work, not just in a centralized place. In that data, through the insights, we can actually improve some of the biggest challenge because I believe climate is a common purpose. Healthcare is a common purpose. Think about dementia, think about Alzheimer. This is an interesting case we are working here in Europe which is basically how we collect the insights from many citizens' brain scans without sharing the actual data and find, figure out how we model the 3D brain to find the protein that ultimately will cure the disease. And we are well on our way, but we call that swarm learning which ultimately is about insights, not about sharing data. So collect all the, I guess if the data is on your balance you probably don't want to share it all, right? Because it has a... But you've shared the insight and you can still create the business out of that. Does anybody else have a philosophy for how they work with China that would be interesting to the room? Ruth, do you want to talk? I should say that when I was at Wired and the Google, the dragonfly stuff came up about Google, there was a question of one Wired employee is it whether Google should be in China and the vote was something like 50 to one, but I was the one saying, I thought that Google should do more in China and push as hard as it possibly could. Do you want to discuss data sharing with your Chinese partners or do you want to... I'm assuming others will have more to say about data sharing with China. Totally fine. That's going out on a limb. I think the question is a red herring, right? So when we run businesses, right, you can sort of do the big bang or you can have progress over perfection. So let's just take the problems that we're talking about, right? The environment. We can make the conversation about China and the rest of the world, or we can say, can we get X number of companies in this industry to work together and not separately on the same topics, whether it's direct air capture and oil and gas or we're working with Eco Patrol, the Columbia oil and gas company, building an open platform around water management because every barrel of oil takes 10 barrels of water. And what Felipe, the CEO, who's very visionary said was this is a big problem for me, but I want to build this. I don't care about the IP. I want it to be an industry solution that we're all implementing because it's not differentiating. Yes, it's going to lower costs. Yes, we can actually do more barrels of oil because of the regulatory constraints, but the reality is if we all come together, share insights, build a more robust platform. Now, I don't need every country in the world. I need the first 10 companies to do it. And so I think the shift that we have to have is how do we stop spending money on things individually? You saw this, the health industry has led it. Clinical trials used to have every different way. Every company did a clinical trial in a different way. They came together under Transcelerate. It's an organization. We have now one way of doing clinical trials. It's used by everyone because it wasn't differentiating to have a different method, but it was a lot of cost, right? And I think that's perfect at the World Economic Forum to start talking about how do we make progress by getting those first 10 or the first three countries? And then as you build it, it's going to make sense to be global. Makes a lot of sense. If I could add on that with respect to risk management. Risk management I think is about having as much horizontal vision as possible. And typically you may just have one small piece of the story, one puzzle piece. And if I think about one of the biggest threats to all of us, it's cyber risk. And at Google, this is one of the core areas, protect our users, protect enterprise customers. But an important group that was brought together a year ago, we were one of the founding members, JCDC. It's the Cyber Defense Collaborative, Joint Cyber Defense Collaborative. And what's key about it is the public sector and the private sector. We each bring whatever puzzle piece we have whenever it emerges. And it's about putting them together as a collective where you then have earlier warning, earlier visibility. And what's important is to continue to expand that, continue to bring in more of the private sector and make it transatlantic. So that we're actually getting the benefit of any kind of threat that we see. We're seeing the same thing, a collective around terrorism, counter-terrorism. And so it's about, again, risk management. We don't each have enough of the story early enough. Well, let's stick on cyber for a second. I think it's one of the most interesting questions. Maybe we'll go to you, Arvind, because I'm watched in one of the high-speed interviews. I believe you said that cyber would be the issue of the next decade. I was having a debate with someone yesterday, and the question was, is the reason Russia's cyber attacks against Ukraine so unsuccessful because, A, they're holding stuff in reserve, or B, because the good guys now have the upper hand on the bad guys, right? Actually, I'm cyber security. Maybe due to this wonderful collaboration, sharing public and private sector, maybe actually the guys and girls trying to stop cyber attacks are doing better than the bad ones. Arvind, what is your take on this? It's always going to be ping-pong back and forth. So the good guys discover better defenses than the Ukraine case. I mean, they obviously knew the Russians would come. By the way, Ukraine should get a lot of credit. They're generally, their cyber skills, not just cyber security, are quite high. And that's been known for a couple of decades. So it's not a surprise to me that they could protect against the Russians coming in. By the way, the simplest way to protect is to disconnect yourself from the network. Leave it local, and we know that's how some of the steel plants are operating. So I'm not surprised that Ukraine being able to defend themselves so far. That said, there is going to be one of these days when a bad actor, probably a nation state, takes down critical infrastructure in some country in a really bad way. That is going to happen. And so it becomes the back and forth. I mean, we talked about the JCDC, but there's other efforts within CISA in the United States, there's other efforts globally. This is going to come about by all of us collaborating on certain things, sharing some insights. I'm not necessarily sharing a quote per se, but also when we patched Linux, for example, with our Red Hat products, those are shared back into the upstream instantly. So you help each other, but it's going to be a ping-pong back and forth. And it's going to be the game is no longer a human game. Till 10 years ago, humans could react. It's going to be a game of machines, and it's going to be on how do you use artificial intelligence, data, and automation. Those are the only three ways to attack this problem. I'll give you a simple example. So we run the website for the masters, which is a golf tournament. So the golf tournament is not important, except you wouldn't expect a lot of bad attacks to come. What are you going to do, take somebody's score? 40 million attacks in less than two days. So now, even if you have a team of 1,000 people, they're not going to pass through a 40 million. So you've got to use automation and AI to pass those down to a few hundred, because then you can have a team of 10 people say, are those serious? That just gives you a scale. I mean, Google, I'm sure, does a lot more. We do about 150 billion events a day for our clients that we have to look at. Without AI automation, those are the techniques to go after. Whereas people think a room full of analysts sitting on monitors, no, no, no. That's maybe for the movies. It's got to be through code, you're going to attack this. So we've got a pretty big portion of the digital economy right here. What is one thing that the five of you should be doing together that you're not doing right now to help us ward off the next terrible cyber attack? Well, if I start, I mean, what you are referring to here is something that we are seeing every day when we are building, I mean, we are building the technology, the network infrastructure technology, the communications networks, both for wireless communications and to fix broadband connections. So I mean, there is a massive increase in interest and requirements from our customers, from operators, from governments, from enterprises. I mean, what we are now doing is that we are building the security, not as an afterthought as it used to be, but it's now going directly into the core, very core of every single product that we make all the way to the chipset level. I mean, you need to have a zero trust attitude to everything in cyber. And even then, you have to assume that bad things will happen. I fully agree with you that it is only a matter of time before we are seeing something going to see. I mean, this Russia, Ukraine, including then Western Europe, I mean, we are all prepared and we are monitoring all the time, but not that much has happened yet, but does that mean that nothing will happen? No, that should not be our attitude. What is the most interesting attack and the most surprising attack? Has somebody tried to build a backdoor in any of your systems? What have you found that's surprising and different that we don't know about? It is really the machines that are attacking in a systematic way, all the different servers, cloud servers, they are attacking our switches, our routers, our radio based stations throughout the network. And we are seeing that the intensity and the, in a way, the intelligence of these attacks is increasing all the time and fully agree. This is not something that human people will be able to solve. I mean, the attackers, they are machines and the defense also needs to be made by machines, not by human. It was unique that 85% of the attacks are now what we consider the core infrastructure, per se. They're coming through sensors, they're coming through fire alarm systems, they're coming through things that you would not imagine because those are now exploding through the part of the growth of IoT. And that's why, yes, we can start with zero trust as a principle by which you build a hardening of the system. But then you have to take care of everything that's connected to the network. And my Google password is 20 letters long, but my Nest password is 1234. So, see why the attacks are coming. So Nick, we can go to the metaverse in terms of what is something new and different that we could all work together that actually the World Economic Forum wants to lead on and that is the responsible metaverse, which the metaverse is in very early stages. And we are talking about, and the WEF wants to help lead, what is a responsible metaverse so that we learn from the past where we had to retrofit a lot around security and the human dimension. So it's around trust, so it's security, data privacy, data rights ownership and identity on the trust side and on the human side. It's about diversity, inclusion, belonging, accessibility. These are areas where we can build into the metaverse now with standards and policies across companies to enable the metaverse to reach its full potential, which there's a ton of potential, which I know you wanted to talk about. Hopefully you've gone to the Global Collaboration Village that we helped build here at the World Economic Forum that shows how you can actually immerse yourself in problems and collaborate differently. That was built in the metaverse. So you had four different companies coming together. We did everything in terms of how the experience would work without ever getting on a plane. But the metaverse itself is at the beginning. And so we can learn from the work that Google has done in trust and safety and the internet and really leading there. All of that can be applied to the metaverse but shouldn't be done. Technology company by technology company, enterprise by enterprise and certainly not government by government. So we're all still trying to get digital standards for responsible AI, right? We should be getting ahead of that with the metaverse. Jose, wonderful note, Julie is promoting my panel tomorrow on setting the rules for the metaverse. Julie, will you explain a little bit about how Accenture uses the metaverse to onboard employees? Because I think it's totally awesome. Absolutely. So 150,000 employees will be onboarded into the Accenture metaverse, which means they will come into the metaverse with peers around the world who are also onboarding on the first day. By the way, if you're trying to tell people that you're an innovative company, getting to onboard in the metaverse is definitely a good way. They will learn about Accenture like our technology. We have something called TQ. Every one of our 700,000 people has to pass and take technology training. They'll learn about what we do. The science says that you actually retain more by 33% more if you are doing it in the metaverse in a 30-minute block, and so we'll have higher retention. But it's also an experience. Our people are still not physically onboarding. And so imagine how you're meeting people all around the world and you're talking about the super cool experience. And that's a very, very new way for us. And it's not simply that we digitized onboarding. We made it completely different. They're immersing in Accenture and meeting each other in a very different way. So I get hired. I lose my job at the Atlantic as media. I get a job at Accenture. You ship me a headset. I put on the headset. I go in. I meet other Accenture employees. Do I have legs in this meeting or do I not have legs? You have arms. I have arms, but no legs. Yes. One of the weird things. So the legless metaverse. So I go into the legless metaverse. I see the Accenture offices. I meet my new colleagues. Well, and you'll go to an innovation hub. So for example, during the pandemic, when we couldn't take clients to innovation hubs to see things, we actually did it in the metaverse. So we'd bring the clients. I'd stand up. I'd give a presentation and oh, by the way, I could now give it in 3D, which I can't do in person. So it's not an alternative. It actually is enhancing how we were immersing. You'd have the same walk around, like have a conference like we do here and have presentations. That's the same experience for our employees come in. They'll meet each other. They'll talk. They'll move around. They'll see a lot of different spaces. They'll be taken to innovation hubs around the world. And that actually is a much bigger improvement than when we pre-pandemic would have them come into the office and we show them pictures, right? We tell them about it. Now they're actually inside the innovation hub. Super cool. All right, let's talk about the industrial metaverse because 24 minutes ago, I said I was going to just have a slight diversion and then bring you back. What is the industrial metaverse? And you have legs. Industrial metaverse is actually an industrial version of what we just heard. So to make it very simple, on the consumer side, I mean, you have all these augmented reality things that we are working on. On the industrial side, it means that first of all, everything that makes sense to connect will be connected. But not only that, there will be a physical world and then there will be a digital world. There will be pretty much a digital twin of everything out there on the industrial side. And what that has mean in practice, it means that a lot of the work that is currently done in the physical world will actually be modeled and sometimes implemented in the digital world. Predictive maintenance has been discussed for a long time. But now, with the help of industrial metaverse, that will get significantly more advanced. We will know when the machines will fail, but not only that, the maintenance technicians will have direct real-time access through the digital world into those machines and they will get trained in that world also. Imagine what would be a good example. A nuclear power plant maintenance engineer. It's a very complicated environment. That whole training onboarding and all of that can be done in the digital world and they will be able to train all kinds of failure situations with the digital twin before they go actually into the physical world. Could it... Okay, so nuclear power plant fails. It's a disaster. You don't want to go in. Radiation is leaking. Not a good situation. But you want to prevent that. Yes, but let's say it happens. Is it conceivable that in the future you will have such a good digital twin that you'll be able to go into the metaverse, into a virtual reality world, turn a dial, pull a switch, feel around and actually change it in the real plant? Absolutely, absolutely. And this is not far off. I mean, the physical and the digital worlds, they will grow together and the result is exactly what you just said. This is knowledge. I mean, like we work with National Grid, which is a utility company. It's dangerous to send people out there. You have a Boston Dynamics robot with our AI on the back end going to fix power lines. One of the LNG terminals in Australia has robots in there because it's dangerous for humans to go in and they go in and they'll turn dials. Now, it is augmented. It's not completely autonomous. So the robot is feeding camera signals, other signals back to an operator or an engineer. They're telling the robot what to do. But the AI is also making recommendations of what to do. So this is now. The same is extremely cool. Okay, let's get... Now, even in the data center space, is augmented reality, we use it all the time to reverse systems. So let's talk a little bit about... Julie mentioned we want to set the rules for the metaverse. We want to set our priorities. We want to get everything right because some people believe that not all the rules have been set exactly correct for the tech sector over the last 10 years or maybe some consequences. What are the choices we need to make now around the metaverse that will affect how it develops in a beneficial way? Anyone can take that. Ruth? Well, if you build off of the work that was done around AI ethical principles, I think you have a starting point. And one of the points that we've spent quite a bit of time on is if you just replicate the world we live in today, we'll recreate and propagate all of the biases that have been built into society. So you need to include humanists, social scientists. You need to bring different perspectives and marry that with computer scientists so that you can actually start to ask yourself, what are the issues that we need to protect? That's one of the core starting points. Great. Another one is really... I mean, we need to get the regulation around this thing right. And there's some worrying development going on, for example, in the European Union at the moment, where, obviously, for very good reasons, there is a strong desire to protect personal data and the way how that is used. But I'm arguing that if we are applying the same artificial intelligence regulation to the data that a turbine in a power plant is producing, we are going to fail. A turbine does not need the same type of personal protection as a human being. And it's very, very important that we get the regulation right. That seems like a winning argument. Are you not able to persuade regulators of that? We've been trying, but sometimes it's harder than you would believe. Identity, the identity of both persons and things. And so the technology is not entirely there, but ultimately, metaverses will be interoperable and you'll want to move your person, your identity, and also your things. You've heard a lot about NFTs right across. So there's a lot of questions around identity. There'll be other technologies like blockchain that'll probably be very critical. We're meeting with Interpol, for example, to think about identity and identity theft. So getting the standards for how you protect the identity, how you actually are able to bring it, and how you protect yourself against fraud and the stealing of identity will be absolutely critical. And that's something that we're gonna need to solve together. I mean, I worry about that if there's gonna be a perfect replica of Nick Thompson in the metaverse, I really don't want that hacked. Somebody else going around creating all these clones of me doing all kinds of terrible things, but I'm glad that we're gonna be able to solve that. Let's move to another big and exciting topic which is quantum computing. I'd love to get all of your sense on where it's going, the problems that we'll solve first, what are the things we need to deal with now? Arvind, you are an expert on this. When are we going to have a real working quantum computer that does something interesting and important that a classical computer cannot solve right now? Yes, that's the last part. So we are working quantum computers now, but they are still inferior to classical computers, so I'll say it that way. 2023, we'll get 1,000 qubits. 50% chance it'll do something. Very quickly, explain what a qubit is. I mean, maybe everybody knows what a qubit is in Stavros, but explain what a qubit is. Quantum bits. So classical computers work on normal bits. So think of a qubit as being somewhat approximately telling you the power of a quantum computer. And is it at 1,000 or 5,000? Somewhere in there, a quantum computer is likely to cross over what anything classical can do, but cross over in a way that probably classical cannot catch up to. So that is why that is an exciting time period. The problems, actually, we just talked about sustainability. So number one, it's going to be able to work on new alloys and new materials. Probably pharmaceuticals are way down the road because there's a bit more complexity in life and death issues, so I'll phrase it that way. Optimization. You talked about the fleet management problem. I think there'll be applications to AI and search, but I think on search, I think Ruth and Google will be far more expert than anything we will do. I think one of the big problems is going to be around climate change. By the way, plankton consume more carbon dioxide than everything else. It's a pretty simple mechanism. But there's zero chance that classical computers are going to figure out how that works. I have a hope that quantum computers can, because you're operating at biochemistry levels. And that's what quantum computers do. They mimic what is happening at the quantum mechanical level. And so there's a chance that they can do problems like nitrogen, carbon dioxide, sequestration, because that, in the end, is a materials problem. Wait, explain what you mean about plankton. Why can't a classical computer model what plankton does? Well, a classical computer, I'll take the caffeine molecule. I'll say the caffeine molecule is a little bit simpler than plankton. A caffeine molecule on a normal classical supercomputer needs a computer the size of this planet. I don't think that's going to happen. So to get all the way to a complete organic mechanism like plankton is impossible. Ruth actually has a computer almost the size of the planet. All right. So quantum computers, what is your prediction for the year when this happens? When we actually have quantum supremacy. Sounds so scary. Commercial advantage. Yeah. I'll go to commercial advantage. Great. When will we have commercial advantage? 2025 would be my prediction. 2025. Not that far. No, that's not that far at all. All right. Ruth, do you agree? You guys are building a quantum computer. We are. And I think the breakthroughs we've had in quantum computing have come probably faster than we even expected, but very much to your point, it's going to take a while before there's commercial application. Like all technologies, they build on themselves. And so we're continuing to learn and accelerate growth. So 2025 is as good a guess as I can come up with. Antonio? But while we work on that, which is so exciting and revolution in many ways, there are also big advances we can drive in the current way we think about conventional computing and think about sustainability along the way. If you think about this room, I mean, problem TV looks big, but it's actually fairly small. As you can see, we have air. And most of the systems today are cooled. But think about liquid way-to-cool systems down, like this glass of water probably is more efficient than the entire room of air. So there is ways to go accelerate the computing aspect as the data continue to outpace computing capacity in a sustainable way. At the same time, you think about the current architecture of computing. It has been for 70 years, Arvin, in the making. And we'll talk about the model's law and talk about how many transistors we can pack on a chip. The fact of the matter is we have to invert the system. We have to think about a data-driven computing, not the CPU-driven computing. And I think that's, in my view, as exciting as quantum computing. Because in the end, you can be way more efficient 1,000 times, way more cheaper in that, and way more sustainable at the same time. So what exactly is the mindset you're describing? It's what we call the memory-driven computing. Basically, if you think about the architecture of a computer, not different than a phone, for the matter, which is a computer. You have a CPU, you have memory and storage. We believe the memory and the storage have to collapse, be persistent, so you need less and less power at a massive scale. And then bring the computing part of it, the CPU, which can be a quantum computing, to the data. So what is the friction? Moving data. Even in your phone, you're moving data all the time. It gets hot, and therefore it consumes energy. We can flip that around. So you're collapsing the memory and the storage into the CPU, and it makes everything more efficient? No, we're collapsing the memory and storage into one single entity and reverse the order by which we compute. That's the key. Oh, I see. OK. And by the way, you don't need gold. You don't need minerals. You can use things like silicon photonics, way cheaper, easy to manufacture on a qubit. Wow. All right. Well, I want to ask one more question about quantum computing, because at the beginning I said there were three conversations that could drive apart any data party. But actually, there's a fourth, which is, should we be scared of quantum computing because it will break encryption? And there are a lot of really smart people who are absolutely convinced that when quantum computing comes, it will destroy all encryption, and all of our services will be hacked. And there are a whole bunch of people who say, oh, come on. That's just not how it works. And I've had really smart people on both sides of this debate. Arvind, you have a view on this. Yeah. Look, you're going to need quantum computers, millions of qubits, and completely error-free before that happens. One. Two, it will break today's encryptions as they're commonly used, because they're based on how do you factor large numbers. And quantum computers at that scale. So I'll come back to when we'll get quantum computers at that scale. The other part, which is what people ought to do, it is plenty of known encryption techniques, which cannot be broken by quantum computers. So go ahead and deploy those. And by the way, certain enterprises already have. A lot of sophisticated enterprises are deploying these techniques. And government standards are working on them, albeit slower than I would like. They could have worked on them a half a dozen years ago, as opposed to waiting till this point. So when are you going to get million qubit quantum computers? That's probably a decade or more away. When are they going to be error-free? That's where there's a big debate. Some people claim never. And some people say. But you don't actually need a quantum computer to be error-free. You just need enough qubit so you can compensate for the errors, right? You begin to blow up in terms of how big it gets. So right now, people say you'll need like 10 or 100 qubits for every qubit if you wanted to sort of emulate something that is error-free. So I think this is at least a decade away. And I'm not worried about it, because if we change now on the encryption standards, then it can't break it. Packer, are you hardening your industrial systems for future quantum attacks? Well, I mean, as I said, security is now inbuilt in everything. And going back to the quantum question, this will be highly relevant for the communications networks as well. Right now, we are all building 5G networks, as we know. But by the time quantum computing is maturing for commercial applications, we are going to be talking about 6G. And 6G will hit the commercial market around 2030. And as I said earlier, we're going to have a digital twin of everything. And that's going to require massive computational resources, including over-the-ear interface. It's a significant computing challenge to be able to transmit all those bits that industrial metaverse will need over the year. We are going to need at least 100 times, if not 1,000 times faster networks over the air. So quantum computing definitely may play a significant role already in the 6G era. The question will be that there's going to be a lot of distributed computing in 6G, small cloud servers here and there, and a lot of intelligence in the radio stations and radio base stations and so on, that will quantum computing scale to decentralized applications as well? Or counterargument, how long will it be relevant only for massive centralized data centers? That's something that, at least to me, is still an unanswered question. Super interesting. My hope, beginning this panel, is that we'd be able to get to a point about 6G quantum attacks on the industrial metaverse. So I'm delighted that we're able to slowly get ourselves there. If there is a tool, there will be a bad guy who will use it. Yeah. So we're going to move to audience questions in just a minute, but I do want to ask a little bit about where we are right now. So maybe I'll go with you, Ruth, which is a lot of fear of a recession, tech sector, all of your stocks are down. Well, not actually all of your stocks are down over the last six months, but most of your stocks are down over the last six months. If here's the recession, you have a fabulous reputation for economic forecasting. Where is the tech sector going to go in the next year? And then we'll talk a little bit about that. And then we'll go to audience questions for the last few minutes. Having lived through a lot of cycles, I think what you see is you come out the back end. And if you look at anyone through the lens rear view mirror, you can see that it's a blip. And that may sound like it's a cop out in answering your question. But I think the things that we've talked about here are going to transform lives for our children and grandchildren. And if we don't keep our eye on them and keep investing in them now, we're going to be missing the opportunity to deliver those breakthroughs. So as long as each company is focused on deep computer science and how that can be transformative, there will be cycles. There will be volatility potentially in the interim. But I think it's our job as leaders of this to look through it and keep investing to deliver. And so short term, long term, we're focused on the long term. Does everybody peacefully agree with that? I saw a lot of smart persons fully agree. Technology is a secular, not a cyclical trend. We will see cycles. And there are worrying science in the air at the moment. None of you CEOs wants to say we should panic and quickly sell everything and move to cash. No, no, no. We just need to be careful. But the good thing is that the technology towards the metaverse, I mean, that is a secular trend. And it will not go away. All right. Let me ask one more question. We'll go to the audience questions. Is there any massive technology trend, you said, that we have to keep investing in the future for our grandparents, that we have not talked about to grandchildren? And our grandpa. Well, they're all going to live to be 150. That's true. We do have to invest for our grandparents. Based on the life sciences panels I've been on here, we're all going to retire at 60 and live to 150. Is there any technology that will be hugely transformative for our grandchildren that we have not been discussing here that we should discuss? Biology and competition brought together. It's going to change how personalized medicine is going to work. I mean, right now, medicine is very statistical, not really personalized except for some very rare cases. So how biology and competition come together and to give Google credit, they did some really good work with their deep mind on the genome sequencing. I think that's a very early example. But I think that's like the very early drop and there's a whole ocean waiting behind it. I think that's going to change how these two sciences come together. And that's gathering momentum. I mean, you're really seeing the shift. There was IT. There was OT operating technology. Now there's ST science technology. And then you add space. So today we are working in space. We're doing space to Earth. So looking at deforestation for clients, we are doing Earth to space. So looking at new materials being built in space. We're even doing some new business models that will anticipate when you actually are doing things in space and commercial models. That's super early. But it is happening today. No, I think it's health care and space to be of the two most revolutionary opportunity for us as a thinker about the future of humanity. Can I add one more? Food production, digital agriculture, virtual farming, digital-enabled farming, totally world revolutionized how we produce food. Wonderful, that is a great list of things. OK, questions from the audience? And if you're behind me, maybe whistle right here, please. Hi, my name? Hi. OK. You're good. Hi, my name's Atoka. I'm a global shaper from Japan. Thank you for the intriguing discussion. So in the discussion, I know you touched on a buzzword metaverse. And there's also Web 3, which is another buzzword, which incorporates decentralization and blockchain technology. And so I'm curious, what are your thoughts on decentralization where data is innately shared? And what are your companies preparing for a decentralized world? Great. Who wants to take that? Well, I mean, I can start. I think some of us grew up, I think, around the table in the centralized world with mainframes and terminals. And we went through this decentralized world with PC and client service. Then we connected the departments and, obviously, the world. And now we live on the centralized world there. This is a grand, beautiful terminal. Can do many things, sometimes detrimental to you. But then, obviously, as we think about the future, we're going to live both in a centralized and decentralized world. I think Peca talked about the concept of moving the cloud and the capabilities of processing data closer when it's generated. In fact, if you think about it, 60% of the data is created where we live, where we work. Right here. We are broadcasting this. Think about how much data we are processing at this time. Sometimes small and a lot. Sometimes big and small quantity. But I think that that has forced us to think about these edge-to-cloud architectures. And the edge is the next frontier in my mind. Think about driving digital transformation starts with the own ramp of secure connectivity. If you are not connected, you are now participating. And that's why the concept of inclusion is so important. The web has been very much at the front to make sure that we treat connectivity as water and electricity. We have to think that way, because it's a resource that everybody has to have access. Obviously, you have Peca here powering a lot of that. So when I think about that, then what is the right model to deploy about analytics? Obviously, Google is doing a lot on that front. And then it goes back to what we discussed, the regulation, the policies, and the ability to share insights, not data. I keep reinforcing that point, to solve some of these big societal problems, whether it's space. We today, by the way, HPE has already landed a supercomputer on the ISS. The astronauts, they don't know anything about computers. They know about research. They think about deforestation. So how we use that data to predict what's coming next. That's a perfect example of decentralized somebody, 240 kilometers above us, is doing processing data there. Can I ask a specific question very much tied to this, which is, do any of you think that more than 10% of your computer systems in five years will be based on blockchain web or web 3M? More than 10%. More than 10%. No. No? No, no, no. We should remain grounded. So it's really important to be future-proofing. It's really important to be doing what Google and IBM are doing on quantum computing. But 90% of the conversations I have with CEOs are they can't even find their data, like let alone talk about decentralized data. I mean, think about the work that Google's doing on data. And I'm like, this is core foundational work that companies are just really beginning. So first of all, they need to get to the cloud. We're only 30% to 40% of the cloud. We call it the formulas, cloud's the enabler. Data will be the driver. And then AI is how you're going to differentiate. But you have to get to the cloud. You need to know where your data is, which is important, and then be able to use AI. That's where most people are, most companies today. That's true, I know. So I just think you have to future-proof it. And that's why you partner with companies like HPE and IBM and Google and Nokia to make sure you've got the foundations right. And you're always thinking ahead. But we're really far from most enterprises in being able to do this work. I got my job as the editor of Wired, that's the only guy I can't even ask you to fix the printer. Right here, please. Hi, I'm Jay Galla. Is it working? Okay, sorry. I'm Jay Galla. I'm a member of parliament from India. And we are currently building a new parliament building. We have 18 official languages in India. And currently in the manual way of doing things, we're only able to have translators translating them into English and Hindi continuously. And Hindi is only spoken by about 35% of the population. So is the translation competent enough now to be able to handle that type of thing through AI into 18 different languages without making mistakes enough because it's their legal and political issues, small changes in translation could create big problems. So... So you just up the stakes on that question at the end. We are excited about the breakthroughs in natural language translation. As I said, we have 150 billion sentences a day that go through Google Translate. And so the breakthrough, the ability to go from this large corpus of data to what we did with English Bengali is what led us to this statement that we would open up 24 additional languages. The quality of the languages is extraordinary. You did up the stakes on this final 0.01%. But we feel very confident about the quality of the languages and we should follow up afterwards. And this is the direction it's going, which is... And it goes back to, I think Julie's point, a lot of what we were talking about now, which is this speed of change does build on itself. And when we see companies moving to the cloud and then figuring out what they're gonna do with data and AI, every time they start, it opens up the next level of analysis. And I'm worried that any company that's behind on application of it, first of all, you do have to get your data organized. We have things like AutoML that help you do it more rapidly. But you're missing on staying connected to customers, operating expenses, risk analytics. So it is moving fast. And I do believe that from a language perspective, people will be wowed at the ability to communicate more broadly. Again, 150 billion sentences a day. And to me, one of the most profound and moving things that I've experienced in the last month is the number of people who've said to me, we were able to communicate with Ukrainian refugees because of Google Translate. It was English to Ukraine, Polish to Ukraine, over and over again. And so it's about connecting the world and to us that's what our mission is. And so a very much an important part of it. And I'm sure you would appreciate with our CEO inspired by the number of languages and the opportunity that work in India. And can I build on Ruth's point? I think that the legal and the regulatory frameworks have to catch up to the technology. With due respect, the human translators probably make more errors. That's fair point. But you didn't sort of acknowledge that. You said, is the technology going to be perfect? Humans are only 80% accurate. So with respect, I think the other frameworks need to catch up. Well, if I can build on that, that I could very much agree. And what we found going back to your point about biology is we had a breakthrough in early detection of metastatic breast cancer. And what we found is that through AI, we were able to see things even better than the best doctors. And will the language translation work as well when they're discussing strict regulation or will it only work when they're discussing deregulation? Well, that's the back door that somebody else is bringing in. All right, we have one minute left, so very quickly. I wanted to ask when you all think we're going to move from this form factor to something that's on your face, glasses and when computing is all on the edge. All right, 50 seconds, who wants to answer quickly? I think it will go, first of all, it will definitely happen. I was talking about 6G earlier, which is around 2030. I would say that by then, definitely the smartphone, as we know it today, will not anymore be the usual, kind of the most common interface. Many of these things will be built directly into our bodies. And one of the important statements we've, I think, made at our most recent developers' conference is that we believe that one of the big advantages of augmented reality is actually solving problems here on Earth. And it will be things like having glasses and being able to translate as you speak with glasses. And those are very close. Holograms will be very, very high quality in the era of 6G. I've already seen 5G-based hologram demonstrations. They already work. But then with the next generation technology, I mean, we could be having this meeting so that in reality, we would be sitting in different parts of the world. All right, well, I've enjoyed seeing you for the last time. I'll ever see you in person. I look forward to seeing you in the next Davos. Thank you very much to our amazing panelists and a wonderful conversation.